We are studying pre-messenger RNA (pre-mRNA) splicing in the budding yeast Saccharomyces cerevisiae. Pre-mRNA splicing takes place in a very large RNA-protein complex, the spliceosome, within which there are subparticles, including the small ribonucleoprotein particles (snRNPs: U1, U2, U4, U5 and U6), each composed of a small nuclear RNA (snRNA) and a set of proteins. The spliceosome is a highly dynamic molecular machine. During spliceosome assembly and during the course of the splicing reactions many dynamic RNA-RNA and RNA-protein conformational changes occur that are regulated by proteins. Proteins also regulate the specificity, accuracy and efficiency of the splicing process. We are investigating the functions of a number of key proteins and characterising their molecular interactions in the spliceosome. In addition, we are using more quantitative systems biology approaches to study the flow of RNA through the various RNA processing pathways and the functional links between these pathways. We use biochemical, cell biological and genetic approaches, including in vitro splicing assays, two-hybrid screens, quantitative real-time RT-PCR, chromatin immunoprecipitation, light microscopy, mass spectrometry, and custom-designed microarrays.

Molecular Interactions of Prp8p

Prp8 protein is a highly conserved, ubiquitously expressed pre-mRNA splicing factor, first identified in yeast. It is a component of spliceosomal U5 snRNPs and is at the catalytic centre of the spliceosome complex in which the splicing reaction occurs. The spliceosome is a highly dynamic RNA-protein complex and Prp8p seems to play a key role in regulating interactions and conformational changes in the catalytic centre. We have studied the interactions of Prp8p with other splicing factors in Saccharomyces cerevisiae. For example the U5 snRNP protein, Brr2p, was shown to bind to both the N-terminal region and the C-terminal region of Prp8p in yeast two hybrid and in vitro experiments (van Nues and Beggs, 2001; Fig. 1).

Prp8p is an exceptionally large protein (280 kDa in yeast). In collaboration with Dr Andrew Newman (MRC Laboratory of Molecular Biology, Cambridge) we have used a genetic dissection approach to identify the domain structure of Prp8p and investigate regions of interactions with other U5 snRNP proteins (Boon et al, 2006; Fig. 1). We are currently studying the protein interactions at the N- and C-terminal regions of Prp8p with the aim of understanding how these plus predicted post-translational modifications may affect conformational changes in the spliceosome. For a review of Prp8p see Grainger and Beggs, 2005.

The importance of Prp8 protein in human disease was realised by McKie et al. (2001) who showed that mutations of several highly conserved residues in the C-terminus of human PRP8 correlated with a severe form of blindness, autosomal dominant Retinitis pigmentosa (RP). Significantly, seven mutations that cause RP in humans, and which affect residues that are identical in the yeast and human proteins, destabilise the C-terminal Prp8p-Brr2p interaction. We obtained evidence for a cytoplasmic precursor U5 snRNP in yeast that lacks Brr2p, and which depends on a nuclear localisation signal in Prp8p for its efficient nuclear import. The association of Brr2p with the U5 snRNP apparently occurs within the nucleus. We found that the introduction of RP mutations in yeast Prp8p result in nuclear accumulation of the precursor U5 snRNP, apparently as a consequence of disrupting the interaction of Prp8p with Brr2p. We therefore proposed a novel assembly pathway for U5 snRNP complexes, that is disrupted by mutations that cause RP in humans (Boon et al., 2007; Fig. 2). In view of the high level of conservation of the U5 snRNP, it seems likely that this novel pathway may be conserved in metazoa and that it has significance for the regulation of U5 snRNP function.

Back to top

Regulation of Brr2p helicase activity

Brr2p is a DExH class RNA helicase required for both spliceosome activation and dissociation of post-splicing complexes (Small et al, 2006 Mol Cell 23:389-399). During spliceosome activation it probably unwinds the U4/U6 heterodimer to permit U2/U6 and 5'SS/U6 interactions. During spliceosome dissociation it may destabilise the interactions among U2, U6 and the products of splicing. Brr2p has 5 identifiable domains, a unique N-terminal domain, two helicase domains (H1 and H2) and two Sec63 domains (Sec63-1 and -2) of unknown function (Fig. 3). Brr2p is unusual in having two helicase domains, but this feature is conserved in Brr2p homologues. Domain H1 is required for Brr2 ATPase activity, whereas two mutations in H2 do not disrupt ATPase activity in vitro. The sequence of the H2 domain is more degenerate, but the predicted secondary structure is highly conserved. We hypothesise that the H2 domain may contribute to Brr2p function in a manner that does not require ATPase activity but it may involve ATP binding. A precedent for this is found in the cystic fibrosis transmembrane conductance regulator (CFTR), a member of the ABC transporter ATPase superfamily. CFTR contains two nucleotide binding domains. Both are required for function, however, only one is catalytically active (Gadsby et al., 2006. Nature 440:477-483). Also, in multimeric DNA helicases, conformational changes regulate the NTPase and/or NTP binding properties at individual subunit interphases. It is postulated that the non-catalytically active sites may play a role in nucleic acid binding or act as regulatory sites (Patel and Picha, 2000. Ann Rev Biochem 69:651-697).

We previously reported that the C-terminal region of Brr2p interacts with several splicing factors, including N-terminal Prp8p (N-Prp8p), Prp2p, Prp16p, Slu7p and Snu66p (van Nues and Beggs, 2001; Fig. 3). We predict that the different activities of Brr2p in spliceosome assembly and disassembly are regulated by interactions with these factors and others yet to be identified. We propose that the H2-Sec63-2 region serves a regulatory rather than a catalytic function through which Brr2p activity is regulated in a temporal manner and through which Brr2p may in turn influence the activity of other factors, including other helicases, conceivably by forming higher order helicase complexes.

Back to top

PRP45: a link between transcription and splicing

Accumulating evidence indicates that, in vivo, spliceosome assembly and splicing occur as the nascent pre-mRNA is being transcribed. This co-transcriptional spliceosome assembly seems to occur in a stepwise fashion following the transcription of the 5'-splice site, the branchpoint sequence and the 3'-splice site. The S. cerevisiae protein Prp45p, a component of the Prp19/Ntc complex was identified by our lab through in vivo interactions with Prp22p and shown to be an essential splicing factor (Albers et al. 2003). The human homologue of Prp45p, SKIP/NCoA-62 has been shown to be both a spliceosome component and a transcriptional co-regulator. Through chromatin immunoprecipitation (ChIP) assays and whole-genome ChIP-chip tiling arrays, we found that, like its human counterpart, Prp45p can interact with the promoter regions of transcribed genes. Furthermore, using a temperature-sensitive allele, prp45-113, we found that growth at the non-permissive temperature causes a transcriptional defect, apparently during the elongation phase. Transcription is subject to checkpoints that involve cycles of phosphorylation and dephosphorylation of serines 2 and 5 in the heptad repeats of the carboxyterminal domain (CTD) of the large subunit of RNA polymerase II (RNAPII; Fig. 4a). The prp45-113 mutation causes a defect in the accumulation of serine 2 phosphorylated RNAPII that could explain the transcription defect. Prp45p interacts both physically and genetically with chromatin remodelling factors, including components of the Isw1b complex that coordinates transcription elongation with termination and RNA processing. We propose that the function of Prp45p is in transcriptional elongation, possibly as a checkpoint factor to couple transcription and splicing in budding yeast (Fig. 4b).

Jean Beggs is also a partner in the EC-funded FP6 Network of Excellence in Alternative Splicing, EURASNET (www.eurasnet.org).

Back to top

Systems Biology

Increasingly, we are applying more systematic approaches, including functional genomic studies using microarray technology and deep sequencing to analyse RNA processing on a genomic scale, and more quantitative and kinetic analyses of RNA processing, using quantitative real-time RT-PCR. Related to this, Jean is coordinator of an EC-funded Framework Programme 6 (FP6) project, "RiboSys", in which we and our collaborators use systems biology approaches to model pre-messenger RNA metabolism in Saccharomyces cerevisiae. Jean is also a partner in the EC-funded FP7 project "UNICELLSYS" and an Associate Director of the Edinburgh Centre for Systems Biology (http://csbe.bio.ed.ac.uk).

Back to top

RiboSys

www.ribosys.org

We are using systems biology approaches to model RNA metabolism in yeast. In order to develop kinetic models, we quantify RNA precursors using quantitative real-time RT-PCR, and determine their rates of production as well as their processing and degradation through the various post-transcriptional pathways. This is complemented by chromatin immunoprecipitation (ChIP) analyses of proteins interacting at the transcribed gene and on the attached nascent transcript. Starting with ab-initio models describing the processing and degradation of yeast pre-messenger RNAs, our modelling partners (Compugen, Israel) are producing mathematical representations and populating the parameters using our quantitative experimental data. Manipulation of the parameters will permit predictions to be made about the behaviour of the systems. These can be tested experimentally using yeast mutants that block specific processing steps. Imaging techniques are being refined and used by our collaborator, Edouard Bertrand (CNRS, Montpellier, France) to visualise individual transcripts to determine whether the population data reflect the situation in individual cells. Our colleague David Tollervey's group, is performing similar studies on the production and processing of pre-ribosomal RNAs. Comparison of the performance of the two models, pre-mRNA and pre-rRNA, should provide further insights and enrich our understanding of both pathways.

Back to top

Unicellsys

www.unicellsys.eu

We are partners in a new EC-funded FP7 project to study the systems biology of control of yeast cell growth and proliferation. The overall objective of UNICELLSYS is a quantitative understanding of fundamental characteristics of eukaryotic unicellular organism biology: how cell growth and proliferation are controlled and coordinated by both extracellular and intrinsic stimuli. Achieving an understanding of the principles with which systems of bio-molecules function requires integrating quantitative experimentation with simulations of dynamic mathematical models in a systems biology approach. We will be analysing RNA metabolism in budding yeast under a variety of growth conditions.

Back to top

The Lsm Proteins

The Lsm (Like Sm) proteins are structurally related to a family of conserved Sm proteins that associate with the snRNAs involved in splicing. We demonstrated that a complex of seven Lsm proteins, Lsm2-8p, accumulates in the nucleus and associates with the U6 snRNA and is important for the stability and/or maturation of the U6 snRNA and for splicing (Cooper et al., 1995; Mayes et al., 1999; Vidal et al., 1999). We showed that Lsm2-8p promotes regeneration of pre-mRNA splicing activity by recycling snRNPs and we proposed that Lsm2-8p complex likely functions as a chaperone to promote conformational rearrangements in RNP particles (Verdone et al. 2004). We showed that the complete Lsm2-8 complex is required for nuclear accumulation of U6 snRNA in yeast (Spiller et al., 2007a). With David Tollervey's group we demonstrated that Lsm proteins are involved in other pathways of RNA metabolism including the degradation of pre-mRNAs and the processing of pre-tRNAs, pre-rRNAs and pre-snoRNAs in the nucleus (Kufel et al, 2002, 2003 & 2004 ) (five papers). For review see Beggs, 2005.

In collaboration with Roy Parker (U. Arizona) we demonstrated that Lsm2p to Lsm7p are also required for decapping of mRNA (Tharun et al., 2000). This led to the finding that the Lsm proteins can form alternative hetero-heptameric complexes, with Lsm1-7p being involved in mRNA decapping in the cytoplasm, and Lsm2-8p being involved in U6 snRNA biogenesis and stability in the nucleus.

We have also analysed the requirements for the assembly of Lsm complexes in vivo and for their distinct cellular localisations. We showed that in yeast, the nuclear accumulation of Lsm proteins depends on complex formation and that the Lsm8p subunit plays a crucial role. We demonstrated that Lsm proteins can change their localisation in response to changes in physiological conditions, such as stress (Spiller et al., 2007b).

P-bodies are cytoplasmic foci that are sites of mRNA degradation and translational repression. The yeast Lsm1-7p complex is recruited to P-bodies under certain stress conditions, and is required for efficient decapping and degradation of mRNAs. We recently showed that the Lsm4p subunit and its asparagine-rich carboxy-terminus are prone to aggregation and that this tendency to aggregate promotes efficient accumulation of Lsm1-7p in P-bodies. The presence of Q/N-rich regions in other P-body components suggests a more general role for aggregation-prone residues in P-body localization and assembly. This is supported by reduced P-body accumulation of other P body components after deletion of such domains (Reijns et al., 2008).

RNOMICS

Edinburgh University was the coordinating partner for a Fifth Framework European Commission project from December 2001 until November 2005. The aims were to identify and characterise links between different pathways of RNA metabolism and improve understanding of how this contributes to the coordination of metabolic processes in a cell. The main processes under study were: pre-mRNA splicing, mRNA 5' end capping, 3' end processing and degradation of mRNAs, pre-mRNAs and small stable RNAs and links between these processes, transcription and translation. Beggs and Tollervey, 2005, describes the general concepts and the approaches used. The partners in the RNOMICS project were Jean Beggs and David Tollervey of Edinburgh University, Alain Jacquier of the Institut Pasteur, Walter Keller of the University of Basel, Maria Carmo-Fonseca of the University of Lisbon and Hybrigenics SA, Paris. For an overview of the results and a list of publications visit ">www.eurnomics.org.

Back to top